13 Confluent Hypergeometric FunctionsKummer Functions

§13.8 Asymptotic Approximations for Large Parameters

Contents
  1. §13.8(i) Large |b|, Fixed a and z
  2. §13.8(ii) Large b and z, Fixed a and b/z
  3. §13.8(iii) Large a
  4. §13.8(iv) Large a and b

§13.8(i) Large |b|, Fixed a and z

If b→∞ in ℂ in such a way that |b+n|≥δ>0 for all n=0,1,2,…, then

For fixed a and z in ℂ

13.8.2 M⁡(a,b,z)∼Γ⁡(b)Γ⁡(b−a)⁢∑s=0∞(a)s⁢qs⁡(z,a)⁢b−s−a,

as b→∞ in |ph⁡b|≤π−δ, where q0⁡(z,a)=1 and

13.8.3 (et−1)a−1⁢exp⁡(t+z⁢(1−e−t))=∑s=0∞qs⁡(z,a)⁢ts+a−1.

When the foregoing results are combined with Kummer’s transformation (13.2.39), an approximation is obtained for the case when |b| is large, and |b−a| and |z| are bounded.

§13.8(ii) Large b and z, Fixed a and b/z

Let λ=z/b>0 and ζ=2⁢(λ−1−ln⁡λ) with sign⁡(ζ)=sign⁡(λ−1). Then

13.8.4 M⁡(a,b,z)∼b12⁢a⁢e14⁢ζ2⁢b⁢(λ⁢(λ−1ζ)a−1⁢U⁡(a−12,−ζ⁢b)+(λ⁢(λ−1ζ)a−1−(ζλ−1)a)⁢U⁡(a−32,−ζ⁢b)ζ⁢b)

and

13.8.5 U⁡(a,b,z)∼b−12⁢a⁢e14⁢ζ2⁢b⁢(λ⁢(λ−1ζ)a−1⁢U⁡(a−12,ζ⁢b)−(λ⁢(λ−1ζ)a−1−(ζλ−1)a)⁢U⁡(a−32,ζ⁢b)ζ⁢b)

as b→∞, uniformly in compact λ-intervals of (0,∞) and compact real a-intervals. For the parabolic cylinder function U see §12.2, and for an extension to an asymptotic expansion see Temme (1978).

Special cases are

13.8.6 M⁡(a,b,b)=π⁢(b2)12⁢a⁢(1Γ⁡(12⁢(a+1))+(a+1)⁢8/b3⁢Γ⁡(12⁢a)+O⁡(1b)),

and

13.8.7 U⁡(a,b,b)=π⁢(2⁢b)−12⁢a⁢(1Γ⁡(12⁢(a+1))−(a+1)⁢8/b3⁢Γ⁡(12⁢a)+O⁡(1b)).

To obtain approximations for M⁡(a,b,z) and U⁡(a,b,z) that hold as b→∞, with a>12−b and z>0 combine (13.14.4), (13.14.5) with §13.20(i).

Also, more complicated—but more powerful—uniform asymptotic approximations can be obtained by combining (13.14.4), (13.14.5) with §§13.20(iii) and 13.20(iv).

For other asymptotic expansions for large b and z see López and Pagola (2010).

For more asymptotic expansions for the cases b→±∞ see Temme (2015, §§10.4 and 22.5)

§13.8(iii) Large a

For the notation see §§10.2(ii), 10.25(ii), and 2.8(iv).

When a→+∞ with b (≤1) fixed,

13.8.8 U⁡(a,b,x)=2⁢e12⁢xΓ⁡(a)⁢(2β⁢tanh⁡(w2)⁢(1−e−wβ)−b⁢β1−b⁢K1−b⁡(2⁢β⁢a)+a−1⁢(a−1+β1+β)1−b⁢e−2⁢β⁢a⁢O⁡(1)),

where w=arccosh⁡(1+(2⁢a)−1⁢x), and β=(w+sinh⁡w)/2. (13.8.8) holds uniformly with respect to x∈[0,∞). For the case b>1 the transformation (13.2.40) can be used.

For an extension to an asymptotic expansion complete with error bounds see Temme (1990b), and for related results see §13.21(i).

When a→−∞ with b (≥1) fixed,

13.8.9 M⁡(a,b,x)=Γ⁡(b)⁢e12⁢x⁢((12⁢b−a)⁢x)12−12⁢b×(Jb−1⁡(2⁢x⁢(b−2⁢a))+env⁡Jb−1⁡(2⁢x⁢(b−2⁢a))⁢O⁡(|a|−12)),

and

13.8.10 U⁡(a,b,x)=Γ⁡(12⁢b−a+12)⁢e12⁢x⁢x12−12⁢b×(cos⁡(a⁢π)⁢Jb−1⁡(2⁢x⁢(b−2⁢a))−sin⁡(a⁢π)⁢Yb−1⁡(2⁢x⁢(b−2⁢a))+env⁡Yb−1⁡(2⁢x⁢(b−2⁢a))⁢O⁡(|a|−12)),

uniformly with respect to bounded positive values of x in each case.

For asymptotic approximations to M⁡(a,b,x) and U⁡(a,b,x) as a→−∞ that hold uniformly with respect to x∈(0,∞) and bounded positive values of (b−1)/|a|, combine (13.14.4), (13.14.5) with §§13.21(ii), 13.21(iii).

When a→∞ in |ph⁡a|≤π−δ and b and z fixed,

13.8.11 U⁡(a,b,z)∼2⁢(z/a)(1−b)/2⁢ez/2Γ⁡(a)⁢(Kb−1⁡(2⁢a⁢z)⁢∑s=0∞ps⁢(z)as+z/a⁢Kb⁡(2⁢a⁢z)⁢∑s=0∞qs⁢(z)as),
13.8.12 𝐌⁡(a,b,z)∼(z/a)(1−b)/2⁢ez/2⁢Γ⁡(1+a−b)Γ⁡(a)×(Ib−1⁡(2⁢a⁢z)⁢∑s=0∞ps⁢(z)as−z/a⁢Ib⁡(2⁢a⁢z)⁢∑s=0∞qs⁢(z)as),
13.8.13 𝐌⁡(−a,b,z)∼(z/a)(1−b)/2⁢ez/2⁢Γ⁡(1+a)Γ⁡(a+b)×(Jb−1⁡(2⁢a⁢z)⁢∑s=0∞ps⁢(z)(−a)s−z/a⁢Jb⁡(2⁢a⁢z)⁢∑s=0∞qs⁢(z)(−a)s),
13.8.14 U⁡(−a,b,z)∼(z/a)(1−b)/2⁢ez/2⁢Γ⁡(1+a)×(Cb−1⁢(a,2⁢a⁢z)⁢∑s=0∞ps⁢(z)(−a)s−z/a⁢Cb⁢(a,2⁢a⁢z)⁢∑s=0∞qs⁢(z)(−a)s),

where Cν⁢(a,ζ)=cos⁡(π⁢a)⁢Jν⁡(ζ)+sin⁡(π⁢a)⁢Yν⁡(ζ) and

13.8.15 pk⁢(z) =∑s=0k(ks)⁢(1−b+s)k−s⁢zs⁢ck+s⁢(z),
qk⁢(z) =∑s=0k(ks)⁢(2−b+s)k−s⁢zs⁢ck+s+1⁢(z)

where c0⁢(z)=1 and

13.8.16 (k+1)⁢ck+1⁢(z)+∑s=0k(b⁢Bs+1(s+1)!+z⁢(s+1)⁢Bs+2(s+2)!)⁢ck−s⁢(z)=0,
k=0,1,2,….

For the Bernoulli numbers Bk see §24.2(i) and for proofs and similar results in which z can also be unbounded see Temme (2015, Chapters 10 and 27)

§13.8(iv) Large a and b

When a,b→+∞ with |z| and ν=ab bounded

13.8.17 M⁡(a,b,z)=eν⁢z⁢Γ∗⁢(b)Γ∗⁢(a)⁢(1+(1−ν)⁢(1+6⁢ν2⁢z2)12⁢a+O⁡(1min⁡(a2,b2))),
13.8.18 U⁡(a,b+1,z)=z−b⁢e(1−ν)⁢z⁢Γ⁡(b)Γ⁡(a)⁢(1+ν⁢z⁢(1−ν)⁢(2−ν⁢z)2⁢a+O⁡(1min⁡(a2,b2))),
ℜ⁡z>0,

where Γ∗⁡(a) is the scaled gamma function defined in (5.11.3). These results follow from Temme (2022), which can also be used to obtain more terms in the expansions. For generalizations in which z is also allowed to be large see Temme and Veling (2022).